Exchange-ification usually improves markets.
Take any OTC market and move it to a central limit order book… spreads tighten, price discovery improves, and counterparty risk evaporates. US rates trading moved from voice brokers to the CME long ago; corporate bonds are undergoing a seismic shift to electronic venues right now; and perps will replace CFDs for simple, leveraged trading on traditional assets very soon.
But prediction markets reveal the limits of this logic.
The Manipulation Paradox
Prediction markets face an impossible problem: the more important they become, the more vulnerable they are to manipulation.
Consider the 2024 US election. Financial markets globally exhibited heightened volatility as asset prices incorporated different probabilities of Trump versus Harris outcomes. Trillions of dollars in assets moved as traders frantically repriced every new piece of information released up to the election.
Now imagine if institutional investors actually used Polymarket or Kalshi prices as serious probability indicators for their portfolio decisions. A manipulator could theoretically move trillions in asset values by deploying a tiny fraction of that capital to aggressively trade a few highly-watched prediction markets.
This isn't theoretical paranoia, it happens every time the market for an underlier (here, the event contract) becomes less liquid than a derivative. When prediction markets become systemically important, they become systemically manipulable. Exit polls, by comparison, cannot be bought.
The Democracy Problem
Prediction market proponents claim superior forecasting ability compared to traditional methods like polling. This argument contains a fundamental category error.
In financial markets, unequal voting power makes sense. When I think Tesla is undervalued, I can buy shares, but my market impact pales compared to Elon Musk's. This asymmetry is rational; Musk has better information about Tesla's fair value and the capital to express that view meaningfully. The market benefits from weighting his opinion more heavily than mine.
But elections operate on a different principle: one person, one vote. Democracy explicitly rejects wealth-weighted decision-making. When prediction markets price election outcomes, they're applying financial market logic to a fundamentally democratic process. To put it explicitly, ten minimum-wage workers have ten times the electoral power of one billionaire, but the billionaire can move prediction market prices more than all ten workers combined. The market is optimizing for the wrong thing.
The Reflexivity Trap
Even worse, I can't make Apple more profitable by buying AAPL shares. But prediction markets often involve participants who can influence the events they're betting on.
Political prediction markets are particularly vulnerable. The billionaire earlier in the thread can simultaneously bet on an election outcome, and fund activities to make that outcome more likely. The prediction becomes self-fulfilling, breaking the information aggregation mechanism that supposedly justifies the market's existence.
This reflexivity problem doesn't exist in traditional financial markets to the same degree. Even Elon Musk can't single-handedly determine Tesla's quarterly earnings by trading Tesla options.
The Great Binary Options Rebrand
The vast majority of the "event contracts" currently marketed in the US are binary options by another name. Binary options were banned in the UK, Europe, and Australia years ago. Not because regulators hate innovation, but because their primary use case was facilitating consumer harm disguised as sophisticated trading.
The exceptions? Betfair and Smarkets in Europe, which have operated exchange-traded gaming on sports and events for over 25 years.
Notice the key word: gaming.
These platforms succeeded by acknowledging what they actually are, rather than cosplaying as financial infrastructure. The inimitable @Agustin Lebrun has a great article on this, but to summarize: real financial markets depend on participants with trading needs other than short-term profits. As a former HFT market maker, I know this all too well; a large part of my job was seeking out this kind of order flow. Indeed, it’s something that has remained consistent throughout the history of leveraged markets, as we discussed in our article here:
(see ARTICLE ON HISTORY OF FUTURES LINK)
Prediction markets lack this natural hedging demand. There's no convincing yes-or-no binary outcome I’ve seen that a corporate treasurer or pension fund would want to hedge against in a binary way, that couldn’t be done better with a regular, linear payoff, with a regular future.
The absence of genuine hedgers means prediction markets can only survive by extracting money from recreational traders - a fundamentally unsustainable model for a financial institution.
Exchange-ification's Limits
I support exchange-ification of fungible contracts because it typically improves price discovery and reduces transaction costs. Moving from bilateral dealer markets to central limit order books benefits nearly everyone except the dealers extracting rent from opacity.
But prediction markets demonstrate that exchange-ification can't solve fundamental structural problems. Better execution and tighter spreads don't address the core issues: manipulation vulnerability, democratic legitimacy, and reflexivity.
Prediction markets remain primarily a vehicle for consumer gambling, regardless of their technological sophistication. Exchange-ification makes them more efficient gambling, not more socially useful.
The Honest Alternative
If we want exchange-traded event betting, let's be honest about what we're building. Betfair's model works because it doesn't pretend to be solving democracy or revolutionizing forecasting. It's a well-regulated gaming platform that happens to use exchange technology.
Financial markets work because they serve genuine economic functions: capital allocation, risk transfer, and price discovery for productive assets. Prediction markets, despite their technological elegance, primarily serve the same function as slot machines: converting human psychology into revenue.
Exchange-ification is a powerful tool, but it can't transform gambling into social utility any more than it can transform lead into gold.
